does price manipulation break standard methods of TA?

[arepo]

SPLIT FROM LUCIF's ANALYSIS THREAD:

All this TA reminds me of voo doo and hocus pocus BS.

If you are here on the forum you ultimately are bullish on bitcoins unless you are working for the FED or other central bank.

I agree- true- the sentiments and psychology of people is the same as in any investment area BUT  in this very small and not very wide spread market we can have a swing of 50% in one day due to a whim of a big investor ( or an early adopter ). So all the sophisticated TA is out the window.

No one person can predict crazy.  And I know some of those early coin holders are probably not all right in the head.

this is a common misconception that has a game theory explanation.

take one assumption: barring 'crazy' actors, TA can accurately account for underlying trends in price data.

hypothesis: while some traders can be considered 'eccentric', direct manipulation is risky and can often incur loss, and a rational actor with a large sum will make 'TA-appropriate' moves (following from our assumption) to maximize gain.

tl;dr -- following a trend has a better profit/risk ratio than manipulating a trend, on average.

--arepo

[arepo]

tl;dr -- following a trend has a better profit/risk ratio than manipulating a trend, on average.

In the fully general case, this is likely correct.

That is: Any fixed strategy to "control" or "manipulate" any market indefinitely will lose, on average, in the long term.

However: If this is the case, why is price manipulation prohibited in mature markets?

I think we can find very good reasons to assume that manipulation events*,
(painting the tape, playing up volume, painting candles near hourly/daily close, decisively breaking resistances, and so on... )
can be very profitable, relative to risk, especially in an immature, unregulated market,
especially one with no underlying asset, especially when trending sideways.

*(Tactical, time-bounded, in pursuit of some definite objective, with a pre-determined failure/stop-loss criterion.)

[arepo]

I think we can find very good reasons to assume that manipulation events*,
(painting the tape, playing up volume, painting candles near hourly/daily close, decisively breaking resistances, and so on... )
can be very profitable, relative to risk, especially in an immature, unregulated market,

the funny thing about regulations is ideally we shouldn't need them, but in practice we often do. maturity is important.

there are two categories of actions you listed: data manipulation, and psychological manipulation.

for data manipulation, i think the sheer cost of such actions should be a sufficient deterrent, as the cost of the manipulation is directly proportional to its magnitude.

as for psychological manipulation, i'm going to get game-theoretical again:

observation: not counting this month, which saw an unprecedented involvement in wet-behind-the-ears traders, the market was visibly maturer than the 'early days' of 2011.

in the case of 'flash crashes' and other sudden, large-volatility movements, the price usually corrects a large percentage of the movement very rapidly, which demonstrates a mature market that can recognize perturbative forces and neutralize them (one manifestation of the self-organizing properties of markets).

in the related case of the breaking of a decisive resistance, they should not be able to reliably make excess profits off of this kind of move because direct manipulation is not possible, since there is no way to know for sure how the market will react to your actions, no matter how large.

as such, i think my hypothesis"following a trend has a better profit/risk ratio than manipulating a trend, on average" can account for the scenarios you described above.

data manipulation decreases profits by incurring proportional costs, and psychological manipulation carries too much risk.

thoughts?

--arepo

[smoothie]

yes

[arepo]

yes

thanks....

[elux]

thoughts?

Thoughts? Oh yes. Will need some time to think, formulate a reply.
Possible keywords: (Stuff I'd like to write and discuss.)

• Markets: Engineers vs. Mathematicians.

• What is manipulation anyway? (Define our terms.)

• Bitcoin's price finding mechanism is an n-player zero-sum game involving money.

• All price-relevant communication is implicitly manipulation, covert or overt.

• Market orders, limit orders are forms of communicating price information. Orders are signaling.

• Market orders are votes. Voting for a trend. Voting against a trend. (Oh, and limit orders are also votes.)

• Psychological tactics vs. technical gambits.

• Technical gambits? (Well, all attempts of manipulation can be reduced to psychology.)

• Can "crazy/wrong" moves profit? Need manipulation be "crazy/wrong"? (Prediction, anticipated response, iterated games.)

• Showing your hand. Implied invisibles. (If I sell 2 000 point zero BTC at market, what, if anything, should that tell you?)

• Imperfect information and choice dissemination of news. (MtGox press releases, market data, etc.)

• Is Bitcoin growing up? (Why these flash-crashes? Why not plain old crashes?)

*sigh*

I wish this was a verbal conversation.

[ElectricMucus]

The only important thing about TA is that it works statistically. If it does it can yield an increase in capital.

[Odalv]

Let's try. If your TA beats this simple table  https://mega.co.nz/#!6cR2kaoC!A-Bk9SSQKBqPj04VieZ9a37xMTpefdP2nvkr8eaCUXY

[seleme]

- what the biggest idiot bull thought is a good price -> High Of The Day.
- what the biggest idiot bear thought is a good price -> Low Of The Day.

me approves this.

[BitcoinAshley]

Standard markets (upon which TA was developed) aren't as susceptible to price manipulation. Remember, paper gold crashed from $15** to $13**, not $15** to $400. And when HFT glitches happen, the market is right back to normal once the bot operators realize what happened. Bitcoin has some very concentrated positions and not much liquidity.

On top of that, it is a true 24 hour global market - this is practically unparalleled. There are few markets that can be said to come close to this. Most global markets are 9-5-ish in each area of the world and some are closed on weekends. So downward price movement does not have 16 hours every single day to cool down in each locale. I am not overestimating the impact of this, I am just stating it as a factor worthy of consideration.

Seeing as TA and its associated patterns/shapes/voodooism was built upon markets with far more liquidity, far more down-time in each locale, and relatively less concentrated positions, not to mention better infrastructure, it makes sense that one would adjust one's TA to allow for far more volatility and general "weird shit." To what extent these factors effectively "break TA"... I have no idea. I guess we'll just have to find out!

Full disclosure: I know absolutely nothing about TA and am just talking out of my ass.   

[ElectricMucus]

just talking out of my ass.   

QFT

[phoenix1]

- what the biggest idiot bull thought is a good price -> High Of The Day.
- what the biggest idiot bear thought is a good price -> Low Of The Day.

 

[wobber]

Let's try. If your TA beats this simple table  https://mega.co.nz/#!6cR2kaoC!A-Bk9SSQKBqPj04VieZ9a37xMTpefdP2nvkr8eaCUXY

What is this??

[Odalv]

Let's try. If your TA beats this simple table  https://mega.co.nz/#!6cR2kaoC!A-Bk9SSQKBqPj04VieZ9a37xMTpefdP2nvkr8eaCUXY

What is this??

Open office  table.  When to sell/buy and how much. No TA needed. :-)

[ajk]

Isnt the answer obviously yes? no amount of TA will work if someone just decides to either buy or sell massive at random

I find these threads funny because a person who uses TA as a main trading strategy  reminds me of the MANY threads I see on poker sites about people asking why they arent winning players, or that they are on Awful downswings

let me guess on why you made this thread, do you make alot more wrong decisions than correct ones? because that is going to happen when you trade basically ALL the time, I make no trades and I made my bitfortune from pennys to a massive amount now and did this without any hassle of daytrading

daytrading is synonymous with gambling hence why you have threads like this asking questions on if TA works or not,
it is NOT an easy life being a day trader because you are going to be wrong probably WAY more than your going to be right, the main idea is to minimize when you lose and maximize when you gain (obviously)

hence why TA can eat crap and why I dont bother,  

[elux]

We need to go deeper. Thought experiment to follow:

Consider the question: "Does technical analysis work, and if so, why?"

To answer this question, we invent a zero-sum n-player game, which we shall call "Bitcoin".

Working from the very simple rules of the game, we can rediscover some pretty cool findings from economics and game-theory.

Note: This is a quick hack first draft. Just laying some groundwork.

The Rules Of The Game are as follows:

• Players compete for points.
• Points have some utility, external to the game.
• Points can be generated at some cost, external to the game.

• Prize pool increases from 0 to 21 million points over the duration of the game.
• Number of points generated per round is cut in half after some constant number of rounds.

• Players are initially uncoordinated.
• Players are allowed to communicate, but compete for points.
• Players tend to seek to maximize their current balance.

• Game state includes points, time, price, volume,.
• Initial conditions are 0, 0, 0, 0, respectively.

Players can perform any one of three actions per round:
• (Buy, Volume).
• (Sell, Volume).
• Abstain.

• Buying comes at some cost, external to the game.
• Selling gives some profit, external to the game.

• Players can make one move per turn.
• Most players will choose to abstain in any given round.
• Some players will make a move, on occasion.

Explaining non-random patterns emerging from the initially random* moves of uncoordinated agents:

• Successful (lucky?) players will try to repeat successful (lucky?) behaviour.

• Successful players have some history of making successful moves.

• Successful players tend to repeat successful moves when similar conditions reappear.
• Successful players tend to communicate their "history".

• Unlucky players tend to shut up, but will try to emulate successful (lucky?) players.

This is sufficient to explain coordination without communication,
as well as how coordination can be augmented by communication.

From whence Technical Analysis?

• When faced with making a decision, players tend to search the game history for similar states, outcomes.

• Given enough data, players will try to find some ([best match],next best-action) rule.
• Human players (involuntarily, unconsciously) invent rules for making predictions.
• Non-human players will be programmed with the formalized but otherwise similar rules to those of human players.

• Players will tend to independently discover the same/similar ([best match],next best-action) in initially random data.
• This causes such patterns to repeat, but only imperfectly. Players try to outwit other players. The game iterates.

• This leads to repeating of historical patterns. Trends emerge.
• Following trends can lead to higher expected score than random actions, or countertrend actions.**

• Technical analysis (various methods to find a next best-action) can be used to coordinate/collaboration of players in the game,
creating/finding Schelling points without need for mutual communication.

*Doubtful, bias likely. (Needs explaining.)
**This statement needs backing, empirical or otherwise.

Wouldn't it be cool to run this as a simulation, with a computer program to implement the game rules, and using humans as players?  

PS: Rediscovering the process that leads to the discovery/creation of hidden Schelling Points feels pretty awesome.

[ManBearPig]

I'm inclined to say no, it's still useful. TA doesn't tell you what to do, simply helps you see what's happened.

[Birdy]

The Rules Of The Game are as follows:
*lots of stuff*

There are some important rules missing:
New players come to the table, old players leave, but the game doesn't stop.
Outside information can influence the game (regulations, media news, critical bugs)
Not all players are aware of all pieces of information.

I doubt you will be able to cover this in a simulation.

[elux]

The Rules Of The Game are as follows:
*lots of stuff*

I doubt you will be able to cover this in a simulation.

Oh, but Bitcoin is that "simulation".

[Odalv]

Let's try. If your TA beats this simple table  https://mega.co.nz/#!6cR2kaoC!A-Bk9SSQKBqPj04VieZ9a37xMTpefdP2nvkr8eaCUXY

What is this??

Open office  table.  When to sell/buy and how much. No TA needed. :-)

When price DOUBLES=200% (or go +30%) SELL 25%(or 10%) of your STASH(e.g.10 BTC)  .. this EXCEL will compute how many BTC have to sell and what will your profit.  .. You can configure it (change red bold numbers)